Description of problem:
since about a week or two, the automatic (or a manual) run of /etc/cron.daily/0logwatch spins up disks that have no mounted partitions and that have been put to sleep with hdparm -y
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. add disk
2. put disk to sleep with eg hdparm -y /dev/sdb
3. run /etc/cron.daily/0logwatch
disk spins up
disk status unaffected
apparently caused by mdadm --examine --scan being invoked in /usr/share/logwatch/scripts/services/mdadm
Workaround: yum downgrade logwatch
It currently downgrades to 7.4.0-18.20120619svn110.fc18.noarch which has no /usr/share/logwatch/scripts/services/mdadm
Filed upstream bug: https://sourceforge.net/tracker/?func=detail&aid=3607400&group_id=312875&atid=1316824
Better workaround: Append the following line to /etc/logwatch/conf/logwatch.conf
Service = "-mdadm"
(In reply to Peter Backes from comment #4)
> Better workaround: Append the following line to
> Service = "-mdadm"
Not a workaround, it's just a matter of configuration. Closing this as NOTABUG.